Localization of the Binding Site for the Oligosaccharide Moiety of Gb

School of Biochemistry and Molecular Biology, UniVersity of Leeds, LS2 9JT, U.K, ... Tufts UniVersity School of Veterinary Medicine, 200 Westboro Road...
0 downloads 0 Views 70KB Size
Biochemistry 2000, 39, 13153-13156

13153

Localization of the Binding Site for the Oligosaccharide Moiety of Gb3 on Verotoxin 1 Using NMR Residual Dipolar Coupling Measurements† Gary S. Thompson,‡ Hiroki Shimizu,‡ Steve W. Homans,*,‡ and Art Donohue-Rolfe§ School of Biochemistry and Molecular Biology, UniVersity of Leeds, LS2 9JT, U.K, and Department of ComparatiVe Medicine, Tufts UniVersity School of Veterinary Medicine, 200 Westboro Road, North Grafton, Massachusetts 01536 ReceiVed June 19, 2000; ReVised Manuscript ReceiVed August 21, 2000

ABSTRACT: By use of NMR residual dipolar coupling measurements in a dilute liquid-crystalline solvent, the solution structure has been determined of the complex between the oligosaccharide moiety of globotriaosylceramide (Gb3-OS) and the B-subunit homopentamer of verotoxin 1 (VTB). The dipolar coupling data indicate that Gb3-OS binds in a single binding site per monomer, which is identical to one of three sites inferred from the X-ray structure of the same complex. We find no evidence within experimental error for occupancy at either of the two additional binding sites observed per monomer in the crystal structure.

The Escherichia coli verotoxins are responsible for microvascular disorders such as haemolytic uremic syndrome (1), which is the principal cause of acute pediatric renal failure. The toxins comprise an enzymatic A subunit in association with a B subunit homopentamer that binds to a specific cell-surface carbohydrate (2-4) which is globotriaosylceramide (Gb3;1 GalR1-4Galβ1-4Glc-Cer) in the case of verotoxin-1. The inhibition of this interaction is the basis of a new proposed therapy (5). The crystal and NMR structures of verotoxin-1 B subunit (VTB) have been solved in the absence (6, 7) and presence of the oligosaccharide moiety of Gb3 (Gb3-OS) (8, 9). The crystal structure of the complex exhibits three binding sites for the glycan. The first, termed site 1, is similar to the site predicted by Nyholm et al. (10, 11) and is characterized by a stacking interaction of Galβ with the side chain of Phe 30. The second site (site 2) is topologically equivalent to the binding sites found in the other OB fold proteins, and is similar to a second site predicted by Nyholm et al. The third site (site 3) involves a stacking interaction of Galβ on the side chain of Trp 34. In our recent study on the solution structure of the complex (9), we found that only site 2 is predominantly occupied in solution. This is supported by a recent study using fluorescence resonance energy transfer (12) which suggests that a coumarin Gb3 analogue binds in a site analogous to site 2, but with a different orientation to that observed in the crystal structure. The transferred nuclear Overhauser effect (TRNOE) techniques that were used in our NMR study are of insufficient accuracy to permit a distinction between the two proposed orientations. Here, we make use of NMR residual † This work was supported by the BBSRC, grants SBD07527 and B06636 * To whom correspondence should be addressed Phone: 0113 233 3125. Fax: 0113 233 3167. E-mail: [email protected]. ‡ University of Leeds. § Tufts University School of Veterinary Medicine. 1 Abbreviations: DHPC, dihexanoylphosphatidylcholine; DMPC, dimyristoylphosphatidylcholine; Gb3, globotriaosylceramide; Gb3-OS, globotriaosylceramide trisaccharide; VTB, verotoxin-1 B subunit.

dipolar coupling measurements on the complex in the weakly aligned state which provides convincing evidence that site 2 is occupied in solution, with an orientation of the glycan that is indistinguishable, within experimental error, from that observed in the crystal. MATERIALS AND METHODS Protein and Ligand Preparation. VTB was expressed and purified as described (13, 14). Uniformly 13C-enriched Gb3OS was prepared from uniformly 13C-enriched glucose by chemical synthesis using a route previously developed in our laboratory (15). Gb3-OS was prepared for NMR studies by dissolution of 0.21 mg. of lyophilised glycan in 603 µL of 7.5% (w/v) dimyristoylphosphatidylcholine/dihexanoylphosphatidylcholine (DMPC/DHPC) (3:1 w/w) in 99.96% D2O doped with 1 mM tetradecyltrimethylammonium bromide (TTAB). Prior to dissolution in the liquid-crystalline medium, the amount of glycan was verified by 1H NMR following dissolution in 99.96% D2O in the presence of 1 mM methanol as standard. For titration studies, the Gb3 in DMPC/DHPC solution was added directly to aliquots of lyophilised VTB (1.16 mg dry weight). The latter were prepared by lyophilisation of a standard solution of VTB whose concentration was determined by optical density (280 1 mg/mL ) 1.073). Following each addition of VTB the degree of magnetic alignment of DMPC/DHPC bicelles was assessed by measurement of the residual quadrupolar splitting of the solvent deuterium resonance (16). Measurement of Residual Dipolar Couplings. One-bond 13C-1H residual dipolar couplings for uniformly 13C-enriched Gb3-OS were measured by use of conventional HSQC experiments in the absence of 13C decoupling in the F2 dimension. Spectral widths were 1200 Hz in the F2 (1H) and 10 kHz in the F1 (13C) dimension, with 4096 and 256 complex datapoints, respectively. Prior to Fourier transformation, data were apodized with cosine-bell weighting functions and zero-filled once, giving a final dataset of 4096 × 256 real datapoints. Residual dipolar couplings for the

10.1021/bi001394+ CCC: $19.00 © 2000 American Chemical Society Published on Web 10/03/2000

13154 Biochemistry, Vol. 39, No. 43, 2000 free ligand were determined from the difference in splittings of the ligand in DMPC/DHPC solution at 25 and 35 °C, corresponding to the isotropic and liquid crystalline phases, respectively. The change in residual dipolar couplings following titration of the ligand with VTB was similarly followed at 35 °C. Data Analysis and Determination of Order Tensor Components. Values of one-bond 13C-1H residual dipolar couplings for the ligand in the bound-state were determined from a plot of fraction of ligand bound versus measured residual dipolar coupling, corrected for the residual dipolar coupling of the free ligand. The former was determined from the known concentrations of VTB and Gb3-OS, together with the Kd for the association (2 × 10-3 M) as reported (17). The residual dipolar couplings for the bound ligand were determined by extrapolation to 100% bound ligand. While in principle there are three binding sites per VTB monomer for Gb3-OS (8), in previous work (9) we showed that the Kds for Gb3-OS at two of these sites are >6 mM. At the ligand:protein ratios used in the present study, the fraction of ligand bound at these lower affinity sites is 1 mM) would be occupied in the crystal. In contrast, a much lower concentration of ligand (0.70 mM) was utilized in our NMR study. With regard to physiological relevance of the binding sites observed it must be borne in mind that the association observed in the present study is monovalent in character, whereas recognition of Gb3 at the

cell surface is likely to be a polyvalent association. It is wellknown that polyvalent associations give rise to greater affinities than the equivalent monovalent association and are generally additive in terms of free energy minus a contribution that Jencks has referred to as the “connection Gibbs energy” (22). Since the association in turn is a logarithmic function of the standard free energy of binding, in practical terms the weaker binding affinities exhibited at sites I and III may contribute to binding under physiological conditions. From the point of view of design of ligand analogues, however, a suitable strategy might be to focus on the highaffinity site. In this regard, it is noteworthy that the STARFISH inhibitor of the Gb3-VTB interaction developed by Bundle and co-workers to bridge sites 1 and 2, in fact binds exclusively to site 2 in two adjacent VTB molecule in the crystal structure (23). The approach described here represents a general method for the delineation of the structures of ligand-protein complexes. In this particular application, the use of 13Cenriched ligand is mandatory, since the protein concentration far exceeds the ligand concentration in the later stages of the titration of Gb3-OS with VTB. However, if lower accuracy in the measured residual dipolar couplings for 100% bound-ligand can be tolerated, the latter can be determined from a single-point measurement at lower protein:ligand ratios. Measurement of heteronuclear residual dipolar couplings then becomes feasible at natural abundance, especially with the advent of cryo-probe technology. REFERENCES 1. Karmali, M. A. (1989) Clin. Microbiol. ReV. 2, 15-38. 2. Samuel, J. E., Perera, L. P., Ward, S., O’Brien, A. D., Ginsburg, V., and Krivan, H. C. (1990) Infect. Immun. 58, 611-618. 3. Waddell, T., Head, S., Petric, M., Cohen, A., and Lingwood, C. (1988) Biochem. Biophys. Res. Commun. 152, 674-679. 4. Lindberg, A. A., Brown, J. E., Stromberg, N., Westlingryd, M., Schultz, J. E., and Karlsson, K. A. (1987) J. Biol. Chem. 262, 1779-1785.

13156 Biochemistry, Vol. 39, No. 43, 2000 5. Armstrong, G. D., Rowe, P. C., Goodyer, P., Orrbine, E., Klassen, T. P., Wells, G., Mackenzie, A., Lior, H., Blanchard, C., Auclair, F., Thompson, B., Rafter, D. J., and McLaine, P. N. (1995) J. Infect. Dis. 171, 1042-1045. 6. Stein, P. E., Boodhoo, A., Tyrrell, G. J., Brunton, J. L., and Read, R. J. (1992) Nature 355, 748-750. 7. Richardson, J. M., Evans, P. D., Homans, S. W., and DonohueRolfe, A. (1997) Nat. Struct. Biol. 4, 190-193. 8. Ling, H., Boodhoo, A., Hazes, B., Cummings, M. D., Armstrong, G. D., Brunton, J. L., and Read, R. J. (1998) Biochemistry 37, 1777-1788. 9. Shimizu, H., Field, R. A., Homans, S. W., and Donohue-Rolfe, A. (1998) Biochemistry 37, 11078-11082. 10. Nyholm, P. G., Brunton, J. L., and Lingwood, C. A. (1995) Int. J. Biol. Macromol. 17, 199-204. 11. Nyholm, P. G., Magnusson, G., Zheng, Z. Z., Norel, R., Binnington-Boyd, B., and Lingwood, C. A. (1996) Chem. Biol. 3, 263-275. 12. Picking, W. D., McCann, J. A., Nutikka, A., and Lingwood, C. A. (1999) Biochemistry 38, 7177-7184. 13. Calderwood, S. B., Acheson, D. W. K., Goldberg, M. B., Boyko, S. A., and Donohue-Rolfe, A. (1990) Infect. Immunol. 58, 2977-2982.

Thompson et al. 14. Acheson, D. W. K., Calderwood, S. B., Boyko, S. A., Lincicome, L. L., Kane, A. V., Donohue-Rolfe, A., and Keusch, G. T. (1993) Infect. Immunol. 61, 1098-1104. 15. Shimizu, H., Brown, J. M., Homans, S. W., and Field, R. A. (1998) Tetrahedron 54, 9489-9506. 16. Tjandra, N., and Bax, A. (1997) Science 278, 1111-1114. 17. St. Hilaire, P. M., Boyd, M. K., and Toone, E. J. (1994) Biochemistry 33, 14452-14463. 18. Losonczi, J. A., Andrec, M., Fischer, M. W. F., and Prestegard, J. H. (1999) J. Magn. Reson. 138, 334-342. 19. Al-Hashimi, H. M., Bolon, P. J., and Prestegard, J. H. (2000) J. Magn. Reson. 142, 153-158. 20. Bru¨nger, A. T. (1993) XPLOR Manual Version 3.1, Yale University, New Haven, CT. 21. Clore, G. M., Gronenborn, A. M., and Tjandra, N. (1998) J. Magn. Reson. 131, 159-162. 22. Jencks, W. P. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 40464050. 23. Kitov, P. I., Sadowska, J. M., Mulvey, G., Armstrong, G. D., Ling, H., Pannu, N. S., Read, R. J., and Bundle, D. R. (2000) Nature 403, 669-672. BI001394+